Using paternity analysis to measure effective pollen dispersal in plant populations

Abstract

Paternity analysis can be used to estimate mean effective pollen dispersal (micro(d)) by sampling offspring from a mother plant and assaying each for a large number of allozyme loci. The male in the population with the highest likelihood of paternity, based entirely on the degree of genetic relationship with the offspring (transition probability) or combined with information on probability of mating with the mother plant, is inferred as the pollen parent. Computer simulations show that the mean distance between inferred males and mother plants (d) reliably estimates micro(d) in defined circumstances. If male mating success decreases with distance from the mother plant, paternity inference based entirely on transition probabilities results in d values that are upwardly biased, perhaps considerably. More reliable estimates can be obtained in this situation when prior information on the general form of the relationship between mating success and distance between mates (the distance function) is used, along with transition probabilities, to infer paternity. However, this procedure is valid only when the general form of the distance function can be reliably assumed. Computer simulations also show that the bootstrap method can be used to closely approximate the SE of .